Subscribe to our 

Pumps & Systems, May 2007

With the recent introduction of a new axial flow pump came a challenge: how to prove to the end user that this pump does, in fact, pump - before it is delivered.

All of the theories, catalogs and performance curves are nice, but nothing beats a "Show Me" proof. Will it pump as much flow as claimed? Will it build up pressure? Will its bearings run cool? Will a seal or packing hold up or leak? As always, our operations manager, Jerry, came up with a solution: just do it!

Testing the pump in a shop is easy, but how do you do it in the field? For ongoing operation, the currently installed pumps are typically running, and production folks frown on any interruptions. For a new installation, there may be little onsite to work with - sometimes, in fact, if the plans and drawings are still in the works there is just an open area and even power hook-ups are not available.

Also, not many plants these days can afford to send their people away from their production facilities to travel somewhere to witness a pump test. "So," suggested Jerry, "if the end user does not come to the test - the test comes to the end user!"

No power to run the motor? No problem, we will bring our generator!

This way, the entire performance curve is produced: flow is measured by a field-transportable ultrasonic flowmeter, head is derived from the pressure gauge readings, and motor power is precalibrated at the shop. The entire head, power and efficiency curves as a function of flow are produced (see the figure below).

The complete performance curve is field-test generated.

But what about the accuracy of flow, head, and power? We must ensure that the instrumentation is properly calibrated to provide accuracy and repeatability consistent with the application requirements, similar to a factory test. With experience and due diligence, head-capacity curves can be generated with enough accuracy to detect even the onset of suction recirculation, abnormal noise, cavitation, and other factors.

Field testing, however, has its own challenges - the main one being accuracy of efficiency. Claims are sometimes made that field testing produces the same accuracy on pump efficiency as shop testing does. These sorts of claims should be taken with caution. Pump efficiency is a ratio of power transmitted to the liquid (fluid power), to the total power (BHP) transmitted to the pump, which includes power losses in bearings, seals, friction, etc., or:

EFF = FHP / BHP

Fluid horsepower, in U.S. units, is:

FHP = Q x H x SG / 3960

Head (H) is the easiest to determine: simply measure the pressures between discharge and suction, with some adjustments for pipe velocities, gage elevations, etc. Flow is a bit more challenging since most field installations do not have flowmeters. Ultrasonic meters that attach to the outside of the pipe are sometimes applied, but typically their accuracy is not always great and depends upon the place they are positioned on the pipe, the nature of the fluid, etc.

The total horsepower can be calculated from measured amps and volts:

BHP = I x V x 1.73 x PFactor x Eff_motor / 746 (1.73 is for the 3-phase motors)

Amps (I) is relatively reliable, and so is voltage (V), but the biggest caveats are power factor and motor efficiency. Power factor typically ranges between 0.88 to 0.92, and motor efficiency between 0.93 to 0.95. However, there is nothing "typical" in the real world. Motor efficiency depends greatly on the motor load itself, and curves depicting this are typically either missing or unavailable. Also, an older motor may be less efficient compared to a new one, even if its curve is available.

Thus, the product of PFactor x Eff_motor is typically assumed as ~0.85, but this assumption carries an error, perhaps as large as ± 5 percent to 10 percent. When all these errors are realized, the efficiency cannot be expected to produce a reliable answer if the objective is to be under 1 percent to 2 percent accuracy. Often the disputed difference between a manufacturer and an end user is even less than that.

So why bother measuring efficiency if it is not accurate? The main reason is trending. Regardless of the error of measurement of any of the variables noted above, use the same (whatever chosen) assumption consistently - today, six months from now, or two years from now. The hard-to-know variables will probably not change that significantly, but you will be able to indirectly identify the condition of your pump by examining a plotted trend.

From trending you can then establish your own guidelines for an overhauling strategy. For example, a 10 percent (gradual) drop in overall efficiency (regardless of absolute accuracy of individual measurement) may indicate worn out rings to a point that the pump needs to be rebuilt. Thus, its overhaul would be planned at a convenient time, such as the next upcoming plant outage.

Field testing such as this also weeds out other issues - hot bearings, assembly errors, leaky seals, etc. In fact, the pumps in our example here come with a reversible gland which adapts to either a mechanical seal or packing arrangement and can be retrofitted in the field.

With proper instrumentation and preparation, a field test can be a powerful and timely tool to ensure the pump will deliver the performance it is supposed to deliver.

Dr. Nelik (aka "Dr. Pump") is president of Pumping Machinery, LLC, an Atlanta-based firm specializing in pump consulting, training, equipment troubleshooting,  pump repairs and engineered upgrades. Dr. Nelik has 30 years experience in pumps and pumping equipment. He has published over fifty documents on pump operations, the engineering aspects of centrifugal and positive displacement pumps, and maintenance methods to improve reliability, increase energy savings, and optimize pump-to-system operations. With questions, comments, or to attend his Pump School, he can be contacted at http://www.pumpingmachinery.com/.

Note: Pumping Machinery has a new line of horizontal and vertically mounted axial flow pumps, including initially (8) sizes. For questions on these products or field testing, contact 770-310-0866 or www.PumpingMachinery.com/products/products.htm.